WO2016115548A1 - Enhanced srs for massive mimo channel estimation - Google Patents
Enhanced srs for massive mimo channel estimation Download PDFInfo
- Publication number
- WO2016115548A1 WO2016115548A1 PCT/US2016/013750 US2016013750W WO2016115548A1 WO 2016115548 A1 WO2016115548 A1 WO 2016115548A1 US 2016013750 W US2016013750 W US 2016013750W WO 2016115548 A1 WO2016115548 A1 WO 2016115548A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- srs
- symbols
- channel
- multiplexing
- ues
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/003—Arrangements for allocating sub-channels of the transmission path
- H04L5/0048—Allocation of pilot signals, i.e. of signals known to the receiver
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
- H04B7/0452—Multi-user MIMO systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
- H04B7/0456—Selection of precoding matrices or codebooks, e.g. using matrices antenna weighting
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
- H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
- H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
- H04B7/0617—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal for beam forming
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
- H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
- H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
- H04B7/0619—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal using feedback from receiving side
- H04B7/0621—Feedback content
- H04B7/0626—Channel coefficients, e.g. channel state information [CSI]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/0001—Arrangements for dividing the transmission path
- H04L5/0003—Two-dimensional division
- H04L5/0005—Time-frequency
- H04L5/0007—Time-frequency the frequencies being orthogonal, e.g. OFDM(A), DMT
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/20—Control channels or signalling for resource management
- H04W72/23—Control channels or signalling for resource management in the downlink direction of a wireless link, i.e. towards a terminal
Definitions
- This invention relates generally to pilot/reference signals used in OFDMA wireless communications for channel estimation and more particularly to the sounding reference signal (SRS) used in LTE uplink and enhanced methods for use in Massive MIMO systems.
- SRS sounding reference signal
- OFDMA wireless communications generally support multiple simultaneous users by scheduling each user a subset of the OFDM frequency band at different points in time.
- the base station (BS) typically sends reference signals distributed across the whole frequen- cy band in every scheduling period, so the user equipment (UE) can do frequent channel estima- tion for the whole frequency band, even though it may not be scheduled to receive any data in that time period or may only be scheduled to receive downlink data in a small portion of that band.
- a UE generally only sends reference signals in the portion of the frequency band allocated for its data transmission.
- LTE provides a sounding reference signal (SRS) mechanism for UEs to send wideband reference signals to allow the BS to do channel estimation for the uplink channel from each UE.
- SRS sounding reference signal
- Base stations with multiple antennas and knowledge of the channel state can use beam- forming to improve system throughput or capacity in bit/s/Hz.
- the SRS channel estimates can be used in constructing the precoding matrix for downlink transmissions and can be used in con- structing the detection matrix for uplink reception. It is required to periodically refresh the chan- nel estimates, since the wireless channel changes over time.
- the channel state is considered to be only valid for providing beamforming gain in the precoding or detection matrix for the "coher- ence time period" and the beamforming gain achievable gradually decreases as time delay after channel estimation increases.
- the coherence time is different for each user within a specific wireless network and depends on UE speed of motion.
- This invention describes SRS transmission methods that are enhanced relative to current LTE Rel.12 SRS transmission methods to meet the Massive MIMO requirements described above.
- These enhancements are described in two separate categories. Firstly, enhancements to multiplex more channels in the same resources for a given SRS transmission and to provide Massive MIMO throughput performance improvements.
- scheduling enhancements to provide adaptive UE-specific scheduling based on channel conditions for each UE. These meth- ods can be combined in different ways in each BS to optimize the channel estimation system ca- pacity for a given level of system performance.
- Fig. 1 shows the time and frequency structure of SRS transmissions, illustrating the method of estimating time and frequency offsets and the different methods of multiplexing many UEs in each RB.
- Fig.2 illustrates the concept of channel separation using the cyclic shift.
- Fig.3 is a flow chart describing the series of algorithm steps for channel separation us- ing cyclic shift.
- Fig. l shows an example of the frequency domain structure of enhanced SRS symbols 1, 2 according to this invention.
- Uplink LTE [1] sends SC-FDMA symbols at a rate of 14000 sym- bols per second.
- Each SC-FDMA symbol contains subcarriers spaced 15 kHz apart arranged in RBs (resource blocks), where each RB has 12 subcarriers per SC-FDMA symbol.
- the number of RBs and so subcarriers used depends on the channel bandwidth as shown in Table 1 below.
- LTE uplink SRS reference signal values for each UE are allocated to these subcarriers. If a conventional SISO or SIMO transmission method was used for SRS, every subcarrier location and every RB would only be allocated to a single UE. LTE provides a MIMO transmission mechanism for SRS, where multiple UEs can be allocated to the same RB with mechanisms that allow channel estimates from each UE to be separated in the BS receiver. With the enhanced SRS signal shown in the Fig. l example, a "cyclic shift" MIMO mechanism is used to multiplex 32 UEs onto the same subcarrier and a "comb" SISO / SIMO mechanism is used to multiplex 2 sets of cyclic- shifted UEs in the same RB. Additionally, not shown in Fig.
- an "orthogonal code" MIMO multiplexing method can be used with N SRS symbols to get a single channel es- timate per subcarrier with a further multiplexing factor of N UEs.
- Cyclic shift MIMO multiplexing mechanism is used in Fig.l to combine 32 UEs onto the same set of subcarriers.
- the principle for separating channel estimates from multiple users within a single SC-FDMA symbol using cyclic shift is shown in Fig.2.
- the reference signal sequence sent by all 32 cyclic shifted is defined by a cyclic shift ⁇ of a base sequence according to
- r(n) is the complex-valued received signal
- s(n) is the known, unity gain
- h(n) is the complex-valued CSI for subcarrier n.
- Dividing the received sample value for each subcarrier by the known CAZAC sequence value is implemented by multiplying r(n) by the complex conjugate value of f U:V (ri). Giving,
- the frequency domain value of any individual resource element, r(n) contains the sum of all channel estimates, which means that it is impossible to separate individual channel estimates directly from r(n) in the frequency domain.
- the IDFT gives all 32 channel impulse responses
- Fig.3 shows the sequence of operations for MIMO channel separation using cyclic shift, i.e.
- a window function or other signal processing function would be used to improve the SNR performance of this procedure.
- Comb SISO / SIMO multiplexing mechanism is used in Fig. l to combine a second set of 32 cyclic- shifted UEs into the same RB as the first set of 32 UEs, using interleaved odd and even samples UE sets. There is no interference between the odd-sample UE set and the even- sample UE set, but the maximum length of each set is now reduced to half of the maximum se- quence length for the LTE bandwidth , so is limited to half the number of RBs de-
- the first 64 UEs defined above multiply their SRS reference signal sequences with orthogonal code [1,1] and the second set of otherwise iden- tical 64 UEs multiply their SRS reference signal sequences with orthogonal code [1,-1]. Adding channel estimates at each subcarrier position across both symbols gives channel estimates for the first set of 64 UEs, the second set cancel to zero (assuming ignorable change in subcarrier chan- nel estimates between the two SRS symbols). Similarly, subtracting channel estimates at each subcarrier position of the 2 nd SRS symbol from the first gives channel estimates for the second set of 64 UEs, the first set cancel to zero.
- the method described here with orthogonal code has the advantage that the aver- age channel estimate per subcarrier for each UE from two SRS symbols is generally more accu- rate than for channel estimates from an individual symbol, the gain from averaging being more than the loss from orthogonal channel code separation.
- the BS has independent frequency domain channel estimates for every transmission path between UE antenna and BS antenna at every subcarrier position trans- mitted.
- this channel estimate is assumed to consist of three independent components, which all need to be estimated accurately to optimize Massive MIMO system performance; the underlying "CSI" (channel state information), "frequency off- set” and "time offset”.
- time and frequency offset are estimated, their effects can be removed from the channel estimates by corresponding de-rotation in time and frequency directions, relative to a reference subcarrier position in time and frequency.
- the lowest subcarrier frequen- cy in the first SC-FDMA SRS symbol can be chosen as the reference.
- de-rotation is ap- plied, the underlying CSI values for each subcarrier are estimated.
- Massive MEVIO performance tests have shown [PPA1], for most channels of interest, that the underlying CSI values, after re- moval of time and frequency offset, do not change substantially between adjacent subcarriers in time and frequency.
- Massive MIMO performance can be obtained by averaging the CSI value over 1 or more RBs and using a single precoding matrix over that set of subcarriers for downlink transmissions and a single detection matrix over that set of subcarriers for uplink transmissions. Where time and frequency offset effects are compensated with the application of a separate diagonal matrix per subcarrier.
- LTE Rel.12 SRS only sends one SRS OFDM symbol at a time. Absolute carrier fre- quency offset between UE and BS cannot be estimated from a single SRS symbol. Enhanced SRS according to this invention can send two adjacent SRS symbols or two symbols close enough to each other in time that the highest frequency of interest can be unambiguously esti- mated as described above.
- Rel.12 LTE [2] supports carrier frequencies up to 3.8GHz with UE frequency accuracy of +/-0. lppm. So, absolute frequency error in carrier frequency in the range +/-380Hz must be tolerated in the BS receiver, in addition to the maximum Doppler shift due to UE motion.
- enhanced SRS in this invention providing two SRS symbols next to each other, or generally in the same TTI, avoids the cost of holding estimated subcarrier values in memory from previous TTI's. Also, supports the estimate including additional Doppler shift without restrictions, including the high speed train channel models in [1]. Provision of multiple SRS symbols per TTI can be provided for LTE according to this invention in multiple ways; For TDD special subframe, according to [1] Table 4.2-1, there are already configurations where two adjacent SRS symbols can be sent in UpPTS sequences of length 4384. Ts.
- the special subframe can be modified to extend the length of UpPTS with corre- sponding decrease in length of DwPTS, to give more SRS transmission opportunities.
- SRS is sent only in the last symbol of the sub- frame. This could be modified, for example, to send optionally in the last two symbols of the subframe or in the last symbol of both slots.
- LTE Rel.12 SRS only allows a maximum of 8 cyclic shifts per symbol.
- Enhanced SRS according to this invention can send as many cyclic shifts as can be practically supported by the cell-specific characteristics of each base station.
- the number of cyclic shifts that can be support- ed in theory, for non-overlapping channel estimates in the time domain, depends on the time de- lay spread of the channel (and maximum time offset).
- the time delay spread of the channel is generally in proportion to the cell size and the approximate cell size is usually known when the BS is configured.
- 10 to 20 time domain samples per cyclic shift window may be enough resolution for good performance, giving correspondingly between 60 and 30 non-overlapping time domain impulse responses and so 60 to 30 UEs multiplexed with cyclic shift and comb of 2.
- Enhanced SRS ac- cording to this invention increases the number of cyclic shifts allowed, supporting many more UEs multiplex on small cells and supporting optimization for the given channel conditions by each individual BS.
- LTE Rel.12 SRS has a fixed comb of 2.
- Enhanced SRS in this invention allows the comb setting to be increased or removed by each base station.
- the combined settings of number of cyclic shifts and comb size can be seen as a single complex optimization problem that can be jointly optimized in the BS to maximize the number of UEs per SRS symbol for a given level of throughput performance in a Massive MEVIO sys- tem, for a given cell configuration.
- Massive MEVIO sys- tem For example, cells with smaller LTE bandwidth have shorter FFT/IFFT length so less resolution in the time domain. Bigger comb size and smaller number of cyclic shifts may work better for a 3MHz carrier than for 20MHz carrier, for a cell in a given lo- cation.
- LTE Rel.12 does not use orthogonal codes for SRS - since it does not support multiple SRS symbols per TTI.
- PUCCH uses orthogonal codes. These could be used to increase the num- ber of UE channel estimates per SRS symbol, for users which already have frequency offset es- timates. The frequency offset will only change very slowly relative to the coherence time of the channel. So once the absolute carrier frequency offset between UE and BS has been calculated from two SRS symbols, on subsequent SRS transmissions, it may be sufficient to do a fine ad- justment, measuring the relative frequency offset from the previous measured value. This could be done with a single SRS channel estimate at each subcarrier for those users. So orthogonal codes in the enhanced SRS in this invention can be used in this way to further increase the num- ber of UEs multiplexed in SRS.
- Estimates of UE speed of motion in the BS PHY can be used to adjust SRS periodicity per user.
- LTE SRS transmissions can be scheduled with user-specific periodicity.
- the rate of up- dating channel estimates for each user for optimal beamforming performance depends on the co- herence time for that user, which correlates with UE speed. The faster the speed, the shorter is the coherence time. Faster UEs would be scheduled more often than slower UEs for optimal sys- tem throughput.
- Extra SRS scheduling opportunities can be allocated in time periods where no uplink data transmissions are scheduled.
- LTE SRS transmissions could optionally be allowed in all up- link subframes and in all symbols. So, for subframes where no UE uplink data transmissions are scheduled, some or all of the symbols in that subframe can be allocated for SRS transmissions. This can be handled with procedures similar to existing uplink scheduling methods. Every UE that sends LTE uplink data in a subframe is previously scheduled to do so by the BS. Where no uplink data is scheduled, those same scheduling resources could be used to schedule SRS with the same latency as for scheduling data.
- a fast SRS scheduling procedure can be added to LTE to provide the most up to date channel estimates for optimal downlink beamforming performance. This would increase down- link latency, since the full cycle of; ⁇ scheduling SRS for a UE, transmitting the SRS scheduling information from BS to UE, detection of scheduling request in the UE, sending SRS from that UE, detection of SRS with channel estimation in BS and precoding matrix calculation ⁇ would all have to complete before downlink data could be sent to that UE. This method would be applied where the quality-of- service requirement for downlink throughput outweighs the requirement for low latency.
- Extra SRS scheduling opportunities can be allocated in the data symbols of downlink TDD subframes where no data is scheduled. This would make use of the fast SRS scheduling mechanism mentioned above or an earlier scheduling period for uplink data, since the LTE scheduling information for downlink data is sent in the control symbols in the same subframe as the downlink data so would arrive too late to schedule uplink SRS in that same subframe.
- a BS with TDD echo canceling capability could allocate uplink SRS transmissions in downlink subframes while it is concurrently sending downlink data. This method would also need to make use of UE location knowledge in the BS to avoid UE to UE interference. The UE sending SRS in a TDD downlink subframe would interfere with a nearby UE receiving its down- link data at the same time.
- SRS symbols can be sent in pairs in the special subframe for TDD to support a new frequency offset calculation.
- Single SRS symbols sent, as in the current LTE FDD standard, can be used to compare with previous stored CSI to update the frequency offset estimate.
- LTE FDD could be modified to send 2 SRS symbols per subframe, to support frequency offset calculation for Massive MIMO.
- One embodiment uses channel separation by cyclic shift, with a configurable number of cyclic shifts that can be optimized for cell size, to maximize the number of users sending pilots on the same resources that can be separated in the time domain. More received pilots per user than possible with dedicated reference subcarriers gives better time offset estimation, critical for Massive MIMO performance.
- Different "orthogonal codes" in UEs with the same cyclic shift and comb parameters can be used to further increase the number of multiplexed users, in cases where a single SRS symbol is sufficient for each UE. For example, where absolute frequency offset has already been calculated and frequency offset measurement relative to the current update is sufficient. .
- Wideband SRS is best for combining most concurrent channels at once, since there is a practical limitation on cyclic shift separation with sequence length.
- reference signals sent along with the data could be used for channel estimation, but generally data resources are allocated for a user for less than the full bandwidth, so generally worse (noisier) channel estimate and fewer concurrent users can be supported than for wideband SRS.
- the detection matrix can be pre-calculated ready for use as soon as the first data symbol arrives for that user.
- UE speed estimation can be used to adjust the frequency of aperiodic SRS transmission, to update CSI for faster moving users more often that slower moving users in the same cell.
- UE's can send more SRS in the guard time. Also, carefully scheduled UEs that don't interfere with each other can send SRS when not re- ceiving.
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Radio Transmission System (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
This invention provides SRS transmission methods to meet the Massive MIMO systems requirements. These methods are described in two separate categories. Firstly, enhancements to multiplex more channels in the same resources for a given SRS transmission and to provide Massive MIMO throughput performance improvements. Secondly, scheduling enhancements to provide adaptive UE-specific scheduling based on channel conditions for each UE. These methods can be combined in different ways in each BS to optimize the channel estimation system capacity for a given level of system performance.
Description
ENHANCED SRS FOR MASSIVE MIMO CHANNEL ESTIMATION
This application claims the benefit of U.S. Provisional Application No. 62/104,639 filed on Jan- uary 16, 2015.
FIELD OF INVENTION
[0001] This invention relates generally to pilot/reference signals used in OFDMA wireless communications for channel estimation and more particularly to the sounding reference signal (SRS) used in LTE uplink and enhanced methods for use in Massive MIMO systems.
BACKGROUND
[0002] OFDMA wireless communications generally support multiple simultaneous users by scheduling each user a subset of the OFDM frequency band at different points in time. In down- link, the base station (BS) typically sends reference signals distributed across the whole frequen- cy band in every scheduling period, so the user equipment (UE) can do frequent channel estima- tion for the whole frequency band, even though it may not be scheduled to receive any data in that time period or may only be scheduled to receive downlink data in a small portion of that band. However, for OFDMA uplink transmissions, a UE generally only sends reference signals in the portion of the frequency band allocated for its data transmission. LTE provides a sounding reference signal (SRS) mechanism for UEs to send wideband reference signals to allow the BS to do channel estimation for the uplink channel from each UE.
[0003] Base stations with multiple antennas and knowledge of the channel state can use beam- forming to improve system throughput or capacity in bit/s/Hz. The SRS channel estimates can be used in constructing the precoding matrix for downlink transmissions and can be used in con- structing the detection matrix for uplink reception. It is required to periodically refresh the chan- nel estimates, since the wireless channel changes over time. The channel state is considered to be only valid for providing beamforming gain in the precoding or detection matrix for the "coher- ence time period" and the beamforming gain achievable gradually decreases as time delay after channel estimation increases. The coherence time is different for each user within a specific wireless network and depends on UE speed of motion.
[0004] Massive MIMO systems increase system data throughput capacity by increasing the
number of users simultaneously sharing the same time and frequency resources in both uplink and downlink data transmission periods. There is a need to correspondingly increase the system capacity for channel estimation from SRS transmissions, in terms of the number of users per SRS transmission and the rate of update per user relative to the coherence time. PCT patent applica- tion PCT/US 15/56500 entitled "Methods for Channel Information Acquisition, Signal Detection and Transmission in Multi-User Wireless Communication Systems" filed on Oct. 20, 2105 (claiming the benefit of U.S. Provisional Application No. 62/065,775, filed on October 20, 2014) has shown that optimal performance in a Massive MIMO system requires perfect knowledge of the following channel characteristics at the instant of applying the precoding or detection matrix; symbol time offset between each UE and the base station, carrier frequency offset between each UE and the base station and the CSI (channel state information) for the multipath fading channel between each UE and BS antenna. Moreover, in a practical system, there is a time delay between calculating these matrices and applying them, so it also follows that for optimal performance, the BS needs perfect knowledge of the UE speed, since this affects the coherence time.
[0005] This invention describes SRS transmission methods that are enhanced relative to current LTE Rel.12 SRS transmission methods to meet the Massive MIMO requirements described above. These enhancements are described in two separate categories. Firstly, enhancements to multiplex more channels in the same resources for a given SRS transmission and to provide Massive MIMO throughput performance improvements. Secondly, scheduling enhancements to provide adaptive UE-specific scheduling based on channel conditions for each UE. These meth- ods can be combined in different ways in each BS to optimize the channel estimation system ca- pacity for a given level of system performance.
BRIEF DESCRIPTION OF DRAWINGS
[0006] Fig. 1 shows the time and frequency structure of SRS transmissions, illustrating the method of estimating time and frequency offsets and the different methods of multiplexing many UEs in each RB.
[0007] Fig.2 illustrates the concept of channel separation using the cyclic shift.
[0008] Fig.3 is a flow chart describing the series of algorithm steps for channel separation us- ing cyclic shift.
DETAILED DESCRIPTION OF THE PRESENT INVENTION
[0009] The detailed description below is intended to highlight the SRS enhancements claimed in this invention relative to the current state of the art SRS used in LTE Rel.12. It is not intended as a full and complete system description. General signal processing techniques and other appli- cations for this invention not mentioned here will be apparent to those skilled in the art. The par- ticular techniques and implementations described are merely exemplary and are not intended to limit the scope of the invention in any way.
[00010] The first category of SRS enhancement, that multiplexes more channels for a given SRS transmission and improves channel estimation performance, is now described.
[00011] Fig. l shows an example of the frequency domain structure of enhanced SRS symbols 1, 2 according to this invention. Uplink LTE [1] sends SC-FDMA symbols at a rate of 14000 sym- bols per second. Each SC-FDMA symbol contains subcarriers spaced 15 kHz apart arranged in RBs (resource blocks), where each RB has 12 subcarriers per SC-FDMA symbol. The number of RBs and so subcarriers used depends on the channel bandwidth as shown in Table 1 below.
[00012] Table 1
The LTE uplink SRS reference signal values for each UE are allocated to these subcarriers. If a conventional SISO or SIMO transmission method was used for SRS, every subcarrier location and every RB would only be allocated to a single UE. LTE provides a MIMO transmission mechanism for SRS, where multiple UEs can be allocated to the same RB with mechanisms that allow channel estimates from each UE to be separated in the BS receiver. With the enhanced SRS signal shown in the Fig. l example, a "cyclic shift" MIMO mechanism is used to multiplex 32 UEs onto the same subcarrier and a "comb" SISO / SIMO mechanism is used to multiplex 2 sets of cyclic- shifted UEs in the same RB. Additionally, not shown in Fig. l, an "orthogonal code" MIMO multiplexing method can be used with N SRS symbols to get a single channel es-
timate per subcarrier with a further multiplexing factor of N UEs. These three multiplexing methods are described below.
[00013] "Cyclic shift" MIMO multiplexing mechanism is used in Fig.l to combine 32 UEs onto the same set of subcarriers. The principle for separating channel estimates from multiple users within a single SC-FDMA symbol using cyclic shift is shown in Fig.2. In accordance with the current LTE standard defined in [1, Sec.5.5.1], the reference signal sequence sent by all 32 cyclic shifted is defined by a cyclic shift α of a base sequence according to
Where, all 32 UEs send the same Constant Amplitude Zero Auto Correlation (CAZAC) base se- quence fU:V(ri) and each UE has a unique value of α. the length of the reference
signal sequence and subcarriers in an RB and is the num-
ber of RBs in an SC_FDMA symbol for the LTE channel bandwidth as defined in Table 1 above. The value of α for the K=32 cyclic shifted UEs in Fig.l is given by
[00014] The ideal channel model for the nth subcarrier in the frequency domain between any one UE transmit antenna and BS receive antenna can be closely approximated as
Where, r(n) is the complex-valued received signal, s(n) is the known, unity gain, complex-valued CAZAC reference signal and h(n) is the complex-valued CSI for subcarrier n. Consider the case where the K=32 cyclic shifted UEs are scheduled to use all RBs in a 20MHz LTE channel. UE0 uses k=0, UE1 uses k=1 etc. After the FFT, the BS received reference signal value in the nth sub- carrier of the SRS signal on a single receive antenna is
Dividing the received sample value for each subcarrier by the known CAZAC sequence value is implemented by multiplying r(n) by the complex conjugate value of fU:V(ri). Giving,
Since for unity gain CAZAC sequence
[00015] So, the frequency domain value of any individual resource element, r(n), contains the sum of all channel estimates, which means that it is impossible to separate individual channel estimates directly from r(n) in the frequency domain.
[00016] Note that the subcarrier values r(n) (in the frequency domain) for any individual chan- nel k, are the frequency response of the channel". From the well-known DFT (discrete
separated in their own time window of samples.
[00017] Fig.3 shows the sequence of operations for MIMO channel separation using cyclic shift, i.e.
1. Correlate the frequency domain received reference signals
where * represents complex conjugate 3
2. Calculate the -point inverse DFT 4
3. Loop, for each channel
domain samples and Set all other time domain samples to zero 6 b) Calculate the to get the frequency domain channel es-
timates for this channel, separated from all other channels 7.
4. End loop. 8
Typically, a window function or other signal processing function would be used to improve the SNR performance of this procedure.
[00018] "Comb" SISO / SIMO multiplexing mechanism is used in Fig. l to combine a second set of 32 cyclic- shifted UEs into the same RB as the first set of 32 UEs, using interleaved odd and even samples UE sets. There is no interference between the odd-sample UE set and the even- sample UE set, but the maximum length of each set is now reduced to half of the maximum se- quence length for the LTE bandwidth , so is limited to half the number of RBs de-
fined for the LTE channel bandwidth in Table 1 above. According to this invention, different values of the comb can be selected, including: 1 (no comb), 2, 3, 4, 6, 12, 24.
[00019] "Orthogonal code" MIMO multiplexing mechanism can also be used with enhanced SRS in this invention to combine channel estimates from multiple, say N, SRS symbols to get a single channel estimate per subcarrier and multiplex N times more users on each SC-FDMA symbol. Consider the two-symbol SRS example shown in Fig. l with 128 UEs transmitting in both SRS symbols. Assume in this example that the first 64 UEs defined above, multiply their SRS reference signal sequences with orthogonal code [1,1] and the second set of otherwise iden- tical 64 UEs multiply their SRS reference signal sequences with orthogonal code [1,-1]. Adding channel estimates at each subcarrier position across both symbols gives channel estimates for the first set of 64 UEs, the second set cancel to zero (assuming ignorable change in subcarrier chan- nel estimates between the two SRS symbols). Similarly, subtracting channel estimates at each subcarrier position of the 2nd SRS symbol from the first gives channel estimates for the second set of 64 UEs, the first set cancel to zero. Although it would be possible to assign a different 64 UEs to each SRS symbol, giving a total of 128 UEs from these two symbols, 64 estimated from each symbol. The method described here with orthogonal code has the advantage that the aver- age channel estimate per subcarrier for each UE from two SRS symbols is generally more accu- rate than for channel estimates from an individual symbol, the gain from averaging being more than the loss from orthogonal channel code separation.
[00020] With the enhanced SRS signal shown in Fig. l, once SEVIO and MIMO signal separation has been applied in the receiver, the BS has independent frequency domain channel estimates for every transmission path between UE antenna and BS antenna at every subcarrier position trans- mitted. For the purposes of the following analysis, this channel estimate is assumed to consist of three independent components, which all need to be estimated accurately to optimize Massive MIMO system performance; the underlying "CSI" (channel state information), "frequency off- set" and "time offset".
[00021] It is known [US689172B 1] that "frequency offset" and "time offset" between each UE transmitter and BS receiver can be estimated from the independent frequency domain channel estimates calculated above, using the principles illustrated in Fig. l. Frequency offset causes a constant phase rotation with time in the frequency domain channel estimates. Correlation be- tween subcarrier values at different OFDM symbols gives a measure of the frequency offset. Time offset causes a constant phase rotation with frequency in the frequency domain channel estimates. Correlation between subcarrier values at different frequencies gives a measure of the
time offset. Fig. l shows frequency offset correlation is calculated between subcarriers at the same frequency, so unbiased by time offset. Similarly, time offset correlation is calculated be- tween subcarriers at the same SC-FDMA symbol, so unbiased by frequency offset. In both cases, the sum of correlations between many subcarriers improves the accuracy of the estimate.
[00022] For frequency offset estimation, the two SC-FDMA symbols used for correlation do not have to be adjacent in time. But the time duration between these symbols defines the maximum frequency offset that can be unambiguously estimated with this method. Where the two SRS symbols in Fig. l are separated in time by sec, frequency offsets up to Hz can be esti-
mated.
[00023] Once time and frequency offset are estimated, their effects can be removed from the channel estimates by corresponding de-rotation in time and frequency directions, relative to a reference subcarrier position in time and frequency. For example, the lowest subcarrier frequen- cy in the first SC-FDMA SRS symbol can be chosen as the reference. Once de-rotation is ap- plied, the underlying CSI values for each subcarrier are estimated. Massive MEVIO performance tests have shown [PPA1], for most channels of interest, that the underlying CSI values, after re- moval of time and frequency offset, do not change substantially between adjacent subcarriers in time and frequency. Moreover, those tests have shown that better Massive MIMO performance can be obtained by averaging the CSI value over 1 or more RBs and using a single precoding matrix over that set of subcarriers for downlink transmissions and a single detection matrix over that set of subcarriers for uplink transmissions. Where time and frequency offset effects are compensated with the application of a separate diagonal matrix per subcarrier.
[00024] The SRS enhancements in this exemplary embodiment compared with state of the art LTE Rel.12 SRS are as follows:
[00025] LTE Rel.12 SRS only sends one SRS OFDM symbol at a time. Absolute carrier fre- quency offset between UE and BS cannot be estimated from a single SRS symbol. Enhanced SRS according to this invention can send two adjacent SRS symbols or two symbols close enough to each other in time that the highest frequency of interest can be unambiguously esti- mated as described above. Rel.12 LTE [2] supports carrier frequencies up to 3.8GHz with UE frequency accuracy of +/-0. lppm. So, absolute frequency error in carrier frequency in the range +/-380Hz must be tolerated in the BS receiver, in addition to the maximum Doppler shift due to UE motion. Although 380Hz frequency offset could be unambiguously measured as described
above with two SRS symbols separated by <= 1/380 = 2.6ms, enhanced SRS in this invention providing two SRS symbols next to each other, or generally in the same TTI, avoids the cost of holding estimated subcarrier values in memory from previous TTI's. Also, supports the estimate including additional Doppler shift without restrictions, including the high speed train channel models in [1]. Provision of multiple SRS symbols per TTI can be provided for LTE according to this invention in multiple ways; For TDD special subframe, according to [1] Table 4.2-1, there are already configurations where two adjacent SRS symbols can be sent in UpPTS sequences of length 4384. Ts. The special subframe can be modified to extend the length of UpPTS with corre- sponding decrease in length of DwPTS, to give more SRS transmission opportunities. For FDD and TDD normal uplink subframes in LTE Rel.12, SRS is sent only in the last symbol of the sub- frame. This could be modified, for example, to send optionally in the last two symbols of the subframe or in the last symbol of both slots.
[00026] LTE Rel.12 SRS only allows a maximum of 8 cyclic shifts per symbol. Enhanced SRS according to this invention can send as many cyclic shifts as can be practically supported by the cell-specific characteristics of each base station. The number of cyclic shifts that can be support- ed in theory, for non-overlapping channel estimates in the time domain, depends on the time de- lay spread of the channel (and maximum time offset). The time delay spread of the channel is generally in proportion to the cell size and the approximate cell size is usually known when the BS is configured.
[00027] For LTE, the largest cell sizes give a time delay spread approximately equal to the cy- clic prefix length. For 20MHz LTE, normal cyclic prefix, the FFT/IFFT has 2048 useful time delay samples per symbol (duration 67μs) and 144 samples of cyclic prefix (duration 4.7μs). 2048/144 = 14.22, giving up to 14 cyclic shifts for SRS.
[00028] Small cell channel models applicable to LTE have significantly smaller time delay spreads. For example, the Indoor Hotspot ITU model in [3] has only 130ns time delay spread. 67000/130 = 515.38, giving, in theory up to 515 cyclic shifts for SRS without overlapping time domain impulse responses. In practice, good performance in cyclic shift channel separation re- quires good resolution of the time domain sampled impulse response. Assume a comb value of 2 as in Fig.l. Then the 600-point inverse DFT in step 2) of the cyclic shift channel separation pro- cedure defined above, with 512 cyclic shifts would only give 600/512=1.17 time domain samples per cyclic shift window. This is not enough resolution for good performance. Though in many
cases, 10 to 20 time domain samples per cyclic shift window may be enough resolution for good performance, giving correspondingly between 60 and 30 non-overlapping time domain impulse responses and so 60 to 30 UEs multiplexed with cyclic shift and comb of 2. Enhanced SRS ac- cording to this invention increases the number of cyclic shifts allowed, supporting many more UEs multiplex on small cells and supporting optimization for the given channel conditions by each individual BS.
[00029] LTE Rel.12 SRS has a fixed comb of 2. Enhanced SRS in this invention allows the comb setting to be increased or removed by each base station.
[00030] The combined settings of number of cyclic shifts and comb size can be seen as a single complex optimization problem that can be jointly optimized in the BS to maximize the number of UEs per SRS symbol for a given level of throughput performance in a Massive MEVIO sys- tem, for a given cell configuration. For example, cells with smaller LTE bandwidth have shorter FFT/IFFT length so less resolution in the time domain. Bigger comb size and smaller number of cyclic shifts may work better for a 3MHz carrier than for 20MHz carrier, for a cell in a given lo- cation.
[00031] LTE Rel.12 does not use orthogonal codes for SRS - since it does not support multiple SRS symbols per TTI. PUCCH uses orthogonal codes. These could be used to increase the num- ber of UE channel estimates per SRS symbol, for users which already have frequency offset es- timates. The frequency offset will only change very slowly relative to the coherence time of the channel. So once the absolute carrier frequency offset between UE and BS has been calculated from two SRS symbols, on subsequent SRS transmissions, it may be sufficient to do a fine ad- justment, measuring the relative frequency offset from the previous measured value. This could be done with a single SRS channel estimate at each subcarrier for those users. So orthogonal codes in the enhanced SRS in this invention can be used in this way to further increase the num- ber of UEs multiplexed in SRS.
[00032] The second category of SRS enhancement in this invention, which provides for adaptive UE-specific scheduling based on channel conditions for each UE, is now described.
[00033] Estimates of UE speed of motion in the BS PHY can be used to adjust SRS periodicity per user. LTE SRS transmissions can be scheduled with user-specific periodicity. The rate of up- dating channel estimates for each user for optimal beamforming performance depends on the co- herence time for that user, which correlates with UE speed. The faster the speed, the shorter is
the coherence time. Faster UEs would be scheduled more often than slower UEs for optimal sys- tem throughput.
[00034] Extra SRS scheduling opportunities can be allocated in time periods where no uplink data transmissions are scheduled. LTE SRS transmissions could optionally be allowed in all up- link subframes and in all symbols. So, for subframes where no UE uplink data transmissions are scheduled, some or all of the symbols in that subframe can be allocated for SRS transmissions. This can be handled with procedures similar to existing uplink scheduling methods. Every UE that sends LTE uplink data in a subframe is previously scheduled to do so by the BS. Where no uplink data is scheduled, those same scheduling resources could be used to schedule SRS with the same latency as for scheduling data.
[00035] A fast SRS scheduling procedure can be added to LTE to provide the most up to date channel estimates for optimal downlink beamforming performance. This would increase down- link latency, since the full cycle of; { scheduling SRS for a UE, transmitting the SRS scheduling information from BS to UE, detection of scheduling request in the UE, sending SRS from that UE, detection of SRS with channel estimation in BS and precoding matrix calculation} would all have to complete before downlink data could be sent to that UE. This method would be applied where the quality-of- service requirement for downlink throughput outweighs the requirement for low latency.
[00036] Extra SRS scheduling opportunities can be allocated in the data symbols of downlink TDD subframes where no data is scheduled. This would make use of the fast SRS scheduling mechanism mentioned above or an earlier scheduling period for uplink data, since the LTE scheduling information for downlink data is sent in the control symbols in the same subframe as the downlink data so would arrive too late to schedule uplink SRS in that same subframe.
[00037] A BS with TDD echo canceling capability could allocate uplink SRS transmissions in downlink subframes while it is concurrently sending downlink data. This method would also need to make use of UE location knowledge in the BS to avoid UE to UE interference. The UE sending SRS in a TDD downlink subframe would interfere with a nearby UE receiving its down- link data at the same time.
[00038] Two symbols of SRS are sent from the same user close together in time to allow fre- quency offset estimation, critical for Massive MIMO performance. SRS symbols can be sent in pairs in the special subframe for TDD to support a new frequency offset calculation. Single SRS
symbols sent, as in the current LTE FDD standard, can be used to compare with previous stored CSI to update the frequency offset estimate. LTE FDD could be modified to send 2 SRS symbols per subframe, to support frequency offset calculation for Massive MIMO.
[00039] One embodiment uses channel separation by cyclic shift, with a configurable number of cyclic shifts that can be optimized for cell size, to maximize the number of users sending pilots on the same resources that can be separated in the time domain. More received pilots per user than possible with dedicated reference subcarriers gives better time offset estimation, critical for Massive MIMO performance. Another method use a configurable "comb" =1, 2 (like SRS), 3, 4, 6, 12, to multiplex more users per SRS symbol.
[00040] Different "orthogonal codes" in UEs with the same cyclic shift and comb parameters can be used to further increase the number of multiplexed users, in cases where a single SRS symbol is sufficient for each UE. For example, where absolute frequency offset has already been calculated and frequency offset measurement relative to the current update is sufficient. .
[00041] Methods of cyclic shift, comb and orthogonal code can be combined in different ways to give each BS total flexibility to optimize; the number of multiplexed SRS users, rate of update of CSI, system throughput performance etc. for a given system configuration and channel condi- tions.
[00042] Wideband SRS is best for combining most concurrent channels at once, since there is a practical limitation on cyclic shift separation with sequence length. For uplink, reference signals sent along with the data could be used for channel estimation, but generally data resources are allocated for a user for less than the full bandwidth, so generally worse (noisier) channel estimate and fewer concurrent users can be supported than for wideband SRS.
[00043] For uplink, lower latency can be achieved using previously sent SRS rather than using reference signals in the data subframes. Using previously sent SRS, the detection matrix can be pre-calculated ready for use as soon as the first data symbol arrives for that user.
[00044] A fast MAC-layer mechanism to request uplink SRS transmission from a user due to be scheduled soon for downlink transmission. For instances where extra latency is acceptable tradeoff for better CSI (so higher throughput) downlink.
[00045] UE speed estimation can be used to adjust the frequency of aperiodic SRS transmission, to update CSI for faster moving users more often that slower moving users in the same cell.
[00046] Schedule extra SRS symbols in uplink subframes whenever bandwidth is available. In LTE, all data sent uplink is known in the BS 4ms ahead of time, since UE's sending uplink data need to be scheduled ahead of time. SRS can be scheduled like data, not needing a fast mecha- nism as in the previous claim.
[00047] For a TDD base station with echo canceller, UE's can send more SRS in the guard time. Also, carefully scheduled UEs that don't interfere with each other can send SRS when not re- ceiving.
[00048] Although the foregoing descriptions of the preferred embodiments of the present inven- tions have shown, described, or illustrated the fundamental novel features or principles of the inventions, it is understood that various omissions, substitutions, and changes in the form of the detail of the methods, elements or apparatuses as illustrated, as well as the uses thereof, may be made by those skilled in the art without departing from the spirit of the present inventions. Hence, the scope of the present inventions should not be limited to the foregoing descriptions. Rather, the principles of the inventions may be applied to a wide range of methods, systems, and apparatuses, to achieve the advantages described herein and to achieve other advantages or to satisfy other objectives as well.
Claims
1. A method for multiplexing a plural of UEs on the same resource element, comprising using SRS transmission to estimate the uplink channel coefficients of each UE; and implementing TO and FO compensation in a MU-MIMO wireless communication system.
served specially for SRS transmission in an uplink subframe.
3. The method in Claim 2 further comprising that the NSRS symbols are preferably continuous when
4. The method in Claim 2 further comprising dividing the Nuse useable subcarriers of a SRS symbol into non-overlap sets or combs, where are mul-
tiplexed on a comb through code division multiplexing (CDM), frequency division multi- plexing (FDM), or a combination of these two multiplexing methods.
comb of the NSRS SRS symbols with the same multiplexing method in each symbol.
configured by the BS according to the application scenario such as cell size, the maximum channel delay spread, the system bandwidth, the number of UEs to be served.
7. The method in Claim 2 further comprising dividing the useable subcarriers of a SRS symbol can into non-overlap sets or combs, where are multi-
plexed on a comb across the NSRS SRS symbols through code division multiplexing (CDM), frequency division multiplexing (FDM), time division multiplexing (TDM), or a combination of the three multiplexing methods.
configured by the BS according to the application scenario such as the cell size, the maxi- mum channel delay spread, the system bandwidth, the number of UEs to be served.
10. The method in Claim 1 further comprising that using the estimates of UE speed of motion in the BS PHY to adjust SRS periodicity per UE.
11. The method in Claim 1 further comprising that extra SRS scheduling opportunities is allocat- ed in time periods where no uplink data transmissions are scheduled besides the specially re- served SRS symbols.
12. The method in Claim 1 further comprising that extra SRS scheduling opportunities is allocat- ed in the data symbols of downlink TDD subframes where no data is scheduled besides the specially reserved SRS symbols.
13. The method in Claim 1 further comprising that a BS with TDD echo canceling capability al- locates uplink SRS transmissions in downlink subframes while it is concurrently sending downlink data.
14. The method in Claim 1 further comprising that the frequency offset of a UE is estimated by comparing the currently estimated CSI with previous stored CSI when single SRS is trans- mitted by a UE.
5. The method in Claim 1 further comprising that the methods of multiplexing UEs, comb and orthogonal sequence are combined in different ways to give each BS the flexibility to opti- mize based on the number of multiplexed SRS users, rate of update of CSI, system through- put performance etc. for a given system configuration and channel conditions.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201680005985.4A CN107210793B (en) | 2015-01-16 | 2016-01-16 | Enhanced SRS for massive MIMO channel estimation |
US15/537,877 US10171217B2 (en) | 2015-01-16 | 2016-01-16 | Enhanced SRS for massive MIMO channel estimation |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US201562104639P | 2015-01-16 | 2015-01-16 | |
US62/104,639 | 2015-01-16 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016115548A1 true WO2016115548A1 (en) | 2016-07-21 |
Family
ID=56406510
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US2016/013750 WO2016115548A1 (en) | 2015-01-16 | 2016-01-16 | Enhanced srs for massive mimo channel estimation |
Country Status (3)
Country | Link |
---|---|
US (1) | US10171217B2 (en) |
CN (1) | CN107210793B (en) |
WO (1) | WO2016115548A1 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108259395A (en) * | 2016-12-29 | 2018-07-06 | 大唐移动通信设备有限公司 | A kind of channel estimation methods and device |
CN109804575A (en) * | 2016-10-11 | 2019-05-24 | 瑞典爱立信有限公司 | The method and apparatus that SRS for adapting in view of measurement process switches |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11363572B2 (en) * | 2016-08-01 | 2022-06-14 | Qualcomm Incorporated | Uplink channel dynamic waveform switching |
EP3688880B1 (en) * | 2017-09-27 | 2021-06-09 | Telefonaktiebolaget LM Ericsson (Publ) | Downlink pilot assignment in massive mimo |
US11382048B2 (en) | 2018-05-22 | 2022-07-05 | Qualcomm Incorporated | Multiplexing solutions in dual connectivity |
CN109104221B (en) * | 2018-06-19 | 2021-04-30 | 南京纳特通信电子有限公司 | Base station test system, method and storage medium based on 3D Massive MIMO |
CN111865450B (en) * | 2019-04-29 | 2022-09-02 | 华为技术有限公司 | Method, device, equipment and storage medium for improving multi-user multiplexing performance |
US11552837B2 (en) * | 2020-01-13 | 2023-01-10 | Samsung Electronics Co., Ltd. | Uplink timing and frequency offset estimation and compensation for CSI estimation and tracking |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2006112030A1 (en) * | 2005-04-14 | 2006-10-26 | Matsushita Electric Industrial Co., Ltd. | Wireless reception apparatus and wireless reception method |
US20100329375A1 (en) * | 2008-01-30 | 2010-12-30 | Kyocera Corporation | Wireless communication system, wireless communication apparatus and wireless communication method |
US20110235602A1 (en) * | 2010-03-25 | 2011-09-29 | Qualcomm Incorporated | Subframe dependent physical uplink control channel (pucch) region design |
US20120063426A1 (en) * | 2009-05-15 | 2012-03-15 | Min Seok Noh | Method and apparatus for transmitting sounding reference signal in radio communication system |
US20120252474A1 (en) * | 2009-11-02 | 2012-10-04 | Esa Tapani Tiirola | Sounding Reference Signal Configuration |
US20120257584A1 (en) * | 2010-02-22 | 2012-10-11 | Telefonaktiebolaget Lm Ericsson (Publ) | Methods and Arrangements in a Telecommunication System |
US20120287900A1 (en) * | 2010-02-04 | 2012-11-15 | Seo Han Byul | Method and device for transmitting a sounding reference signal |
US20130343359A1 (en) * | 2008-02-05 | 2013-12-26 | Samsung Electronics Co., Ltd. | Method and apparatus for transmitting srs in lte tdd system |
US20140126464A1 (en) * | 2011-07-08 | 2014-05-08 | Sca Ipla Holdings Inc. | Mobile communications network, mobile communications device, relay node and method |
WO2014077742A1 (en) * | 2012-11-13 | 2014-05-22 | Telefonaktiebolaget L M Ericsson (Publ) | Methods and apparatus for reference signal antenna mapping configuration |
US20140241242A1 (en) * | 2013-02-27 | 2014-08-28 | Samsung Electronics Co., Ltd | Methods and apparatus for channel sounding in beamformed massive mimo systems |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011166699A (en) * | 2010-02-15 | 2011-08-25 | Ntt Docomo Inc | Radio base station device, mobile terminal device, and radio communication method |
WO2013119073A1 (en) * | 2012-02-11 | 2013-08-15 | 엘지전자 주식회사 | Method for reporting channel state information, method for supporting same, and apparatus for said methods |
US9503925B2 (en) * | 2012-07-23 | 2016-11-22 | Lg Electronics Inc. | Method for measuring and reporting CSI-RS in wireless communication system, and apparatus for supporting same |
-
2016
- 2016-01-16 CN CN201680005985.4A patent/CN107210793B/en active Active
- 2016-01-16 WO PCT/US2016/013750 patent/WO2016115548A1/en active Application Filing
- 2016-01-16 US US15/537,877 patent/US10171217B2/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2006112030A1 (en) * | 2005-04-14 | 2006-10-26 | Matsushita Electric Industrial Co., Ltd. | Wireless reception apparatus and wireless reception method |
US20100329375A1 (en) * | 2008-01-30 | 2010-12-30 | Kyocera Corporation | Wireless communication system, wireless communication apparatus and wireless communication method |
US20130343359A1 (en) * | 2008-02-05 | 2013-12-26 | Samsung Electronics Co., Ltd. | Method and apparatus for transmitting srs in lte tdd system |
US20120063426A1 (en) * | 2009-05-15 | 2012-03-15 | Min Seok Noh | Method and apparatus for transmitting sounding reference signal in radio communication system |
US20120252474A1 (en) * | 2009-11-02 | 2012-10-04 | Esa Tapani Tiirola | Sounding Reference Signal Configuration |
US20120287900A1 (en) * | 2010-02-04 | 2012-11-15 | Seo Han Byul | Method and device for transmitting a sounding reference signal |
US20120257584A1 (en) * | 2010-02-22 | 2012-10-11 | Telefonaktiebolaget Lm Ericsson (Publ) | Methods and Arrangements in a Telecommunication System |
US20110235602A1 (en) * | 2010-03-25 | 2011-09-29 | Qualcomm Incorporated | Subframe dependent physical uplink control channel (pucch) region design |
US20140126464A1 (en) * | 2011-07-08 | 2014-05-08 | Sca Ipla Holdings Inc. | Mobile communications network, mobile communications device, relay node and method |
WO2014077742A1 (en) * | 2012-11-13 | 2014-05-22 | Telefonaktiebolaget L M Ericsson (Publ) | Methods and apparatus for reference signal antenna mapping configuration |
US20140241242A1 (en) * | 2013-02-27 | 2014-08-28 | Samsung Electronics Co., Ltd | Methods and apparatus for channel sounding in beamformed massive mimo systems |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109804575A (en) * | 2016-10-11 | 2019-05-24 | 瑞典爱立信有限公司 | The method and apparatus that SRS for adapting in view of measurement process switches |
US11239930B2 (en) | 2016-10-11 | 2022-02-01 | Telefonaktiebolaget Lm Ericsson (Publ) | Methods and apparatus for adapting SRS switching accounting for measurement procedure |
CN108259395A (en) * | 2016-12-29 | 2018-07-06 | 大唐移动通信设备有限公司 | A kind of channel estimation methods and device |
EP3565199A4 (en) * | 2016-12-29 | 2019-12-18 | Datang Mobile Communications Equipment Co., Ltd. | Channel estimation method and device |
CN108259395B (en) * | 2016-12-29 | 2021-01-22 | 大唐移动通信设备有限公司 | Channel estimation method and device |
US10999105B2 (en) | 2016-12-29 | 2021-05-04 | Datang Mobile Communications Equipment Co., Ltd | Channel estimation method and device |
Also Published As
Publication number | Publication date |
---|---|
US20180026765A1 (en) | 2018-01-25 |
CN107210793B (en) | 2020-07-24 |
US10171217B2 (en) | 2019-01-01 |
CN107210793A (en) | 2017-09-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10171217B2 (en) | Enhanced SRS for massive MIMO channel estimation | |
US20240137257A1 (en) | Sounding reference signal processing for lte | |
CN107005320B (en) | For channel information acquisition, signal detection and the method for transmission in multi-user wireless communication system | |
CN109478970B (en) | Method and apparatus for transmitting DMRS in wireless communication system | |
US9288020B2 (en) | Transmission of sounding reference signals in TDD communication systems | |
US7848438B2 (en) | Method and apparatus for pilot signal transmission | |
RU2369030C2 (en) | Modulating system with pluralty of carriers with cyclic delays diversity | |
JP5663163B2 (en) | Method, terminal and uplink system for determining uplink demodulated pilot sequence | |
EP2157717B1 (en) | Method and system and device for CQI estimation disturbance cancellation | |
CN114946162B (en) | Ambiguity/aliasing cancellation or reduction for interleaved reference signals | |
CN107210808B (en) | Method for estimating uplink SINR and method for estimating channel error level | |
CA2794068A1 (en) | Method for sending uplink sounding reference signal, method for estimating channel, mobile terminal, base station and wireless communication system | |
KR101181774B1 (en) | Method for allocating pilot signal in multi-carrier system | |
Li et al. | On channel estimation for multi-user MIMO in LTE-A uplink | |
KR20110093555A (en) | Method and apparatus for transmitting csi-rs according to feedback of mobile terminal | |
RU2487477C1 (en) | Method and apparatus for transmitting correction code in wireless communication system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 16738027 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 01.12.2017) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 16738027 Country of ref document: EP Kind code of ref document: A1 |